人本视角街道绿视率与鸟瞰视角绿化覆盖率的表现差异及影响因素解析  

Mapping the Differences Between Human-Scaled Visible Street Greenery vs.Green Coverage from Bird's-Eye View and a Further Exploration of the Effecting Issues

作　　者：叶宇[1,2] 黄成成[1,3] 李心恬 陈晓雨 YE Yu;HUANG Chengcheng;LI Xintian;CHEN Xiaoyu(College of Architecture and Urban Planning(CAUP), Tongji University;Key Laboratory of Ecology and Energy-saving Study of Dense Habitat, Ministry of Education, Tongji University;Computational Urban Design Lab, Tongji University;“Joint International Research Laboratory of Eco-urban Design, Ministry of Education”, affiliated with the College of Architecture and Urban Planning(CAUP), Tongji University)

机构地区：[1]同济大学建筑与城市规划学院 [2]同济大学高密度人居环境与生态节能教育部重点实验室 [3]同济大学计算性城市设计实验室 [4]同济大学建筑与城市规划学院教育部生态化城市设计国际合作联合实验室

出　　处：《风景园林》2023年第9期20-28,共9页Landscape Architecture

基　　金：国家自然科学基金“基于多源数据和深度学习的公共空间品质评价模型与设计支持研究”(编号52078343);上海市自然科学基金面上项目“多源数据与城市计算支持下的市民生活便利度测度模型研究”(编号20ZR1462200);中央高校基本科研业务费专项资金“多源数据与城市计算支持下的市民生活便利度评价模型研究”(编号22120210540)。

摘　　要：【目的】在绿色城市设计兴起的背景下,人本视角街道绿视率作为城市空间精细化感知品质的指征日益受到重视。探索人本视角街道绿视率与现行规划管控所使用的鸟瞰视角绿化覆盖率之间的关系,以揭示绿化覆盖率是否能够充分反映市民在日常生活中频繁接触的街道绿视率水平,旨在为将街道绿视率指标纳入绿色城市设计导控提供科学依据。【方法】借助街景大数据与卫星遥感影像,运用深度学习与地理信息系统,以定性的四象限分析、定量的逻辑回归分析和相关性分析探索中国8个城市的街道绿视率与绿化覆盖率的一致性表现情况。【结果】发现一线、新一线城市的街道绿视率和绿化覆盖率往往具有一致性,而二线城市大概率不一致。街道绿视率与绿化覆盖率的一致性表现,除受自然气候条件的影响,还受经济水平的显著正向影响;街道绿视率除受绿化覆盖率和经济水平的正向影响,还受街块面积的负向影响。【结论】街道绿视率与绿化覆盖率的一致性表现并非必然,有必要将街道绿视率作为导控要素纳入绿色城市设计中进行分析。街道绿视率与绿化覆盖率的一致性,以及街道绿视率指标自身的高低并不单纯由自然气候条件决定,适度的财政投入能有效提升街道绿视率与绿化覆盖率,小街密路的城市形态特征则能有效提升街道绿视率。[Objective] The relationship between visible street greenery from the humanistic perspective(an indicator of refined perceptual quality) and green coverage from the bird's-eye view that is commonly employed in planning and control practice is explored to provide theoretical and practical support for the effective incorporation of visible street greenery into green urban design and planning.[Methods] Data are collected,and green coverage,visible street greenery,and potential influencing factors are calculated.Each sample's green coverage is estimated using satellite images and GIS to calculate their vegetation cover with the normalized difference vegetation index(NDVI).Additionally,deep learning algorithms are used to extract green areas from streetscape big data and calculate them through GIS.Urban morphology data are obtained by processing big data through ArcGIS,natural conditions data are obtained from www.tianqi.com,and economic-level data are obtained from relevant statistical yearbooks.Moreover,the correlation between visible street greenery and greenery coverage in different areas is discussed by a four-quadrant classification method:consistent visible street greenery and green coverage(visible street greenery and green coverage are both high or both low,category A);high green coverage while low visible street greenery(category B);high visible street greenery while low green coverage(category C).A four-quadrant multiple logistic regression is used to analyze the impact of potential influencing factors on the formation of the aforesaid three categories of relationships between visible street greenery and green coverage.Finally,multiple linear regression is used to analyze the influence of each potential factor on visible street greenery.[Results] The economic level may influence the performance of both visible street greenery and green coverage.Generally,first-tier and new first-tier cities exhibited consistent visible street greenery and greenery coverage,while second-tier cities displayed inconsistencies.

关 键 词：绿色城市设计 街道绿视率 绿化覆盖率 多源数据 

分 类 号：TU985.12[建筑科学—城市规划与设计]